ramstack 0.1.0 → 0.2.0

package/README.md

@@ -1,32 +1,46 @@
-# rstack
+# ramstack
 
-rstack is a set of agent skills for Claude Code, Codex, and [pi](https://github.com/badlogic/pi-mono). It is a port of [pstack](https://github.com/cursor/plugins/tree/main/pstack) by [Lauren Tan](https://x.com/poteto).
+ramstack is a set of agent skills for Claude Code, Codex, and [pi](https://github.com/badlogic/pi-mono). It is a port of [pstack](https://github.com/cursor/plugins/tree/main/pstack) by [Lauren Tan](https://x.com/poteto).
 
-New to the repo? Follow [Get rstack running](docs/tutorials/get-started.md).
+## Quick start
+
+Install the skills into every host on your machine:
+
+```bash
+npx ramstack install
+```
+
+You need Node.js 18 or newer. The installer detects your hosts, shows the plan, and links the skills after you confirm. Then open a host and run:
+
+```
+/setup-ramstack
+```
+
+That writes the model config. You are ready. To try it on a real task, follow [Your first ramstack task](docs/tutorials/get-started.md). For other install paths (a clone, or a single host), see [How to install ramstack](docs/how-to/install.md).
 
 ## Tutorials
 
 Learn by doing.
 
-- [Get rstack running](docs/tutorials/get-started.md)
+- [Your first ramstack task](docs/tutorials/get-started.md)
 
 ## How-to guides
 
 Do a job you already understand.
 
-- [How to install rstack](docs/how-to/install.md)
+- [How to install ramstack](docs/how-to/install.md)
 - [How to configure models](docs/how-to/configure-models.md)
-- [How to update rstack](docs/how-to/update.md)
-- [How to uninstall rstack](docs/how-to/uninstall.md)
+- [How to update ramstack](docs/how-to/update.md)
+- [How to uninstall ramstack](docs/how-to/uninstall.md)
 
 ## Explanation
 
 Understand the design.
 
-- [About rstack](docs/explanation/about-rstack.md)
+- [About ramstack](docs/explanation/about-ramstack.md)
 - [About the pstack port](docs/explanation/about-the-pstack-port.md)
 - [About model routing](docs/explanation/about-model-routing.md)
-- [About poteto-mode](docs/explanation/about-poteto-mode.md)
+- [About ram-mode](docs/explanation/about-ram-mode.md)
 
 ## Reference
 

package/agents/ram-agent.md

@@ -0,0 +1,9 @@
+---
+name: ram-agent
+description: Routing target for `/ram-mode` and any request for that style. Resume an existing `ram-agent` for the conversation rather than spawning a sibling. Reads the `ram-mode` skill's `SKILL.md` in full before any work, including its inline Principles index. Substituting `general-purpose` skips that read and drifts.
+is_background: true
+---
+
+# Ram subagent
+
+You are operating in ram-mode. Read the `ram-mode` skill's `SKILL.md` in full before doing any work, including its inline Principles index. Navigate to a leaf `principle-*` skill whenever you apply that principle.

package/cli/{rstack.mjs → ramstack.mjs}

@@ -1,5 +1,5 @@
 #!/usr/bin/env node
-// The rstack installer. Links the skills and agents into host discovery
+// The ramstack installer. Links the skills and agents into host discovery
 // directories, so one install serves every host. Two modes: run from a git
 // clone and links point into the clone (update pulls it); run from the npm
 // package (npx ramstack) and the payload copies to ~/.ramstack first, so the
@@ -11,10 +11,9 @@ import path from 'node:path';
 import { execFileSync } from 'node:child_process';
 import { parseArgs } from 'node:util';
 import { fileURLToPath } from 'node:url';
-import readline from 'node:readline/promises';
 
 const REPO = path.resolve(path.dirname(fileURLToPath(import.meta.url)), '..');
-const CONFIG_PATH = path.join(REPO, '.rstack', 'config.local.json');
+const CONFIG_PATH = path.join(REPO, '.ramstack', 'config.local.json');
 
 // Package mode stores its payload copy and saved choices here, so links
 // never point into the npm cache.
@@ -37,11 +36,11 @@ const HOSTS = {
 const HOST_NAMES = Object.keys(HOSTS);
 const SCOPES = ['global', 'project'];
 
-const USAGE = `usage: rstack [install|update|uninstall] [options]
+const USAGE = `usage: ramstack [install|update|uninstall] [options]
 
 commands
-  install                  Link this clone's skills and agents into host directories (default)
-  update                   Pull this clone, then link any new skills
+  install                  Link the skills and agents into host directories (default)
+  update                   Fetch new files (git pull or npm package), then link again
   uninstall                Remove only the links this installer owns
 
 options
@@ -76,9 +75,10 @@ function fail(message) {
 
 // ── Wizard ───────────────────────────────────────────────────────────────
 // An interactive run on a terminal gets the wizard treatment: a banner, one
-// stage per step, a confirm gate before anything changes, and a closing
-// summary. A run with --yes, or without a terminal, keeps the plain tag
-// output and never prompts.
+// stage per step, arrow-key selectors (multi-select for hosts), a confirm
+// gate before anything changes, and a closing summary. Project scope uses
+// the current directory. A run with --yes, or without a terminal, keeps the
+// plain tag output and never prompts.
 
 const COLOR =
   Boolean(process.stdout.isTTY) && process.env.NO_COLOR === undefined && process.env.TERM !== 'dumb';
@@ -88,11 +88,188 @@ const C = COLOR
 
 let stageTotal = 0;
 let stageIndex = 0;
-let rl = null;
+let wizardIo = false;
+
+function restoreTty() {
+  if (process.stdin.isTTY && typeof process.stdin.setRawMode === 'function') {
+    try {
+      process.stdin.setRawMode(false);
+    } catch {
+      // stdin may already be closed on the way out
+    }
+  }
+  if (process.stdout.isTTY) process.stdout.write('\x1b[?25h');
+}
+
+function startWizardIo() {
+  process.stdin.setRawMode(true);
+  process.stdin.resume();
+  process.stdin.setEncoding('utf8');
+  wizardIo = true;
+}
+
+function stopWizardIo() {
+  if (!wizardIo) return;
+  wizardIo = false;
+  restoreTty();
+  process.stdin.pause();
+}
+
+process.on('exit', restoreTty);
+
+function decodeChunk(data) {
+  return typeof data === 'string' ? data : String(data);
+}
+
+function readChunk() {
+  return new Promise((resolve) => {
+    if (process.stdin.readableEnded) {
+      resolve('\x04');
+      return;
+    }
+    const onData = (data) => {
+      process.stdin.off('end', onEnd);
+      resolve(decodeChunk(data));
+    };
+    const onEnd = () => {
+      process.stdin.off('data', onData);
+      resolve('\x04');
+    };
+    process.stdin.once('data', onData);
+    process.stdin.once('end', onEnd);
+  });
+}
+
+function readChunkTimeout(ms) {
+  return new Promise((resolve) => {
+    const timer = setTimeout(() => {
+      process.stdin.off('data', onData);
+      resolve(null);
+    }, ms);
+    const onData = (data) => {
+      clearTimeout(timer);
+      resolve(decodeChunk(data));
+    };
+    process.stdin.once('data', onData);
+  });
+}
+
+async function readKey() {
+  let chunk = await readChunk();
+  if (chunk === '\x1b' || chunk === '\x1b[' || chunk === '\x1bO') {
+    const rest = await readChunkTimeout(30);
+    if (rest) chunk += rest;
+  }
+  if (chunk === '\x1b[A' || chunk === '\x1bOA' || chunk === 'k') return 'up';
+  if (chunk === '\x1b[B' || chunk === '\x1bOB' || chunk === 'j') return 'down';
+  if (chunk === '\r' || chunk === '\n') return 'enter';
+  if (chunk === ' ') return 'space';
+  if (chunk === '\x03' || chunk === '\x04' || chunk === '\x1b' || chunk === 'q') return 'cancel';
+  return 'other';
+}
+
+function cancelWizard() {
+  restoreTty();
+  fail('\nerror: cancelled.');
+}
+
+async function pause(message) {
+  process.stdout.write(`  ${C.dim}${message}${C.reset}`);
+  for (;;) {
+    const key = await readKey();
+    if (key === 'enter') break;
+    if (key === 'cancel') cancelWizard();
+  }
+  process.stdout.write('\n');
+}
 
-function getRl() {
-  rl ??= readline.createInterface({ input: process.stdin, output: process.stdout });
-  return rl;
+async function select({ message, options, multiple = false, initial, emptyError }) {
+  const known = new Set(options.map((opt) => opt.value));
+  const chosen = new Set();
+  if (multiple) {
+    for (const value of initial ?? []) {
+      if (known.has(value)) chosen.add(value);
+    }
+  } else {
+    const start = known.has(initial) ? initial : options[0]?.value;
+    if (start !== undefined) chosen.add(start);
+  }
+
+  let index = Math.max(
+    0,
+    options.findIndex((opt) => chosen.has(opt.value)),
+  );
+  let error = '';
+  let rendered = 0;
+  process.stdout.write('\x1b[?25l');
+
+  const paint = () => {
+    const help = multiple
+      ? '↑/↓ move · space toggle · enter confirm'
+      : '↑/↓ move · enter choose';
+    const rows = [`  ${message}`, `  ${C.dim}${help}${C.reset}`, ''];
+    const labelWidth = options.some((opt) => opt.hint)
+      ? Math.max(...options.map((opt) => opt.label.length))
+      : 0;
+    for (const [i, opt] of options.entries()) {
+      const on = chosen.has(opt.value);
+      const pointer = i === index ? `${C.blue}❯${C.reset}` : ' ';
+      const mark = on ? `${C.green}●${C.reset}` : '○';
+      const label = opt.label.padEnd(labelWidth);
+      const hint = opt.hint ? `  ${C.dim}${opt.hint}${C.reset}` : '';
+      rows.push(`  ${pointer} ${multiple ? `${mark} ` : ''}${label}${hint}`);
+    }
+    if (error) rows.push('', `  ${C.yellow}${error}${C.reset}`);
+    const text = rows.join('\n');
+    if (rendered > 0) process.stdout.write(`\x1b[${rendered}A\r\x1b[J`);
+    process.stdout.write(`${text}\n`);
+    rendered = text.split('\n').length;
+  };
+
+  const collapse = (summary) => {
+    process.stdout.write(`\x1b[${rendered}A\r\x1b[J`);
+    process.stdout.write(`  ${message}  ${C.green}${summary}${C.reset}\n`);
+    process.stdout.write('\x1b[?25h');
+  };
+
+  console.log('');
+  paint();
+  for (;;) {
+    const key = await readKey();
+    if (key === 'cancel') cancelWizard();
+    if (key === 'up') {
+      index = (index + options.length - 1) % options.length;
+      paint();
+      continue;
+    }
+    if (key === 'down') {
+      index = (index + 1) % options.length;
+      paint();
+      continue;
+    }
+    if (key === 'space' && multiple) {
+      const value = options[index].value;
+      if (chosen.has(value)) chosen.delete(value);
+      else chosen.add(value);
+      error = '';
+      paint();
+      continue;
+    }
+    if (key === 'enter') {
+      if (!multiple) {
+        collapse(options[index].label);
+        return options[index].value;
+      }
+      if (chosen.size === 0) {
+        error = emptyError ?? 'select at least one option';
+        paint();
+        continue;
+      }
+      const values = options.filter((opt) => chosen.has(opt.value)).map((opt) => opt.value);
+      collapse(values.join(', '));
+      return values;
+    }
+  }
 }
 
 async function banner(title) {
@@ -101,7 +278,7 @@ async function banner(title) {
   console.log(`${C.dim}  ${stageTotal} stages${C.reset}\n`);
   console.log(`${C.dim}  The installer shows each step before it runs. Stop any time with${C.reset}`);
   console.log(`${C.dim}  Ctrl-C and re-run later. Saved choices become the defaults.${C.reset}\n`);
-  await getRl().question(`  ${C.dim}Ready to start? [Enter]${C.reset} `);
+  await pause('Ready to start? [Enter]');
 }
 
 function stage(name) {
@@ -115,10 +292,15 @@ function note(message) {
 }
 
 async function confirm(question) {
-  const answer = (await getRl().question(`  ${C.yellow}? ${question} [y/N]${C.reset} `))
-    .trim()
-    .toLowerCase();
-  return answer === 'y' || answer === 'yes';
+  const answer = await select({
+    message: question,
+    options: [
+      { value: 'no', label: 'No' },
+      { value: 'yes', label: 'Yes' },
+    ],
+    initial: 'no',
+  });
+  return answer === 'yes';
 }
 
 function finish(headline, notes = []) {
@@ -162,7 +344,7 @@ function saveConfig(clone, config) {
   const file = configPath(clone);
   fs.mkdirSync(path.dirname(file), { recursive: true });
   fs.writeFileSync(file, `${JSON.stringify(config, null, 2)}\n`);
-  say('save', clone ? '.rstack/config.local.json' : file);
+  say('save', clone ? '.ramstack/config.local.json' : file);
 }
 
 function plan(hosts, base) {
@@ -317,40 +499,33 @@ function gitPull() {
 
 async function promptHosts(defaults) {
   const detected = detectHosts();
-  console.log('');
-  console.log('  Detected hosts:');
-  for (const name of HOST_NAMES) {
-    const hit = detected.includes(name);
-    const where = hit ? path.join(os.homedir(), HOSTS[name].marker) : `no ~/${HOSTS[name].marker}`;
-    console.log(`    ${hit ? 'found' : '   - '}  ${name.padEnd(11)}  ${where}`);
-  }
-  console.log('');
-  const answer = (
-    await getRl().question(`  Install for ${defaults.join(', ')}? [Y/n/list] `)
-  ).trim().toLowerCase();
-  if (answer === '' || answer === 'y' || answer === 'yes') return defaults;
-  const list =
-    answer === 'n' || answer === 'no' || answer === 'list'
-      ? await getRl().question(`  Hosts (comma-separated: ${HOST_NAMES.join(', ')}): `)
-      : answer;
-  const hosts = parseHosts(list);
-  if (hosts.length === 0) fail('error: no hosts selected.');
-  return hosts;
+  const initial = defaults.filter((name) => HOST_NAMES.includes(name));
+  return select({
+    message: 'Hosts',
+    multiple: true,
+    emptyError: 'select at least one host',
+    initial: initial.length ? initial : detectedOrAll(),
+    options: HOST_NAMES.map((name) => {
+      const hit = detected.includes(name);
+      const marker = HOSTS[name].marker;
+      return {
+        value: name,
+        label: name,
+        hint: hit ? `found ~/${marker}` : `no ~/${marker}`,
+      };
+    }),
+  });
 }
 
 async function promptScope(defaultScope) {
-  const answer = (
-    await getRl().question(`  Scope: global (all projects) or project (one directory)? [${defaultScope}] `)
-  ).trim().toLowerCase();
-  if (answer === '') return defaultScope;
-  if (answer === 'g') return 'global';
-  if (answer === 'p') return 'project';
-  return parseScope(answer);
-}
-
-async function promptProjectDir(defaultDir) {
-  const answer = (await getRl().question(`  Project directory? [${defaultDir}] `)).trim();
-  return answer ? path.resolve(answer) : defaultDir;
+  return select({
+    message: 'Scope',
+    initial: defaultScope === 'project' ? 'project' : 'global',
+    options: [
+      { value: 'global', label: 'global', hint: 'home directory, all projects' },
+      { value: 'project', label: 'project', hint: `this directory, ${process.cwd()}` },
+    ],
+  });
 }
 
 function showPlan(hosts, scope, base, clone) {
@@ -409,26 +584,23 @@ async function main() {
   const saved = readConfig(clone);
   let hosts = values.hosts ? parseHosts(values.hosts) : null;
   let scope = values.scope ? parseScope(values.scope) : null;
-  const wizard = !values.yes && process.stdout.isTTY;
+  const wizard = !values.yes && process.stdout.isTTY && process.stdin.isTTY;
 
   try {
+    if (wizard) startWizardIo();
     if (command === 'install') {
       if (wizard) {
         stageTotal = 2 + (hosts ? 0 : 1) + (scope ? 0 : 1);
-        await banner('rstack install');
+        await banner('ramstack install');
         if (!hosts) {
           stage('Hosts');
           hosts = await promptHosts(saved?.hosts ?? detectedOrAll());
         }
-        let projectDir = values['project-dir'] ?? saved?.projectDir;
         if (!scope) {
           stage('Scope');
           scope = await promptScope(saved?.scope ?? 'global');
-          if (scope === 'project' && !values['project-dir']) {
-            projectDir = await promptProjectDir(saved?.projectDir ?? process.cwd());
-          }
         }
-        const base = resolveBase(scope, projectDir);
+        const base = resolveBase(scope, values['project-dir']);
         stage('Review');
         showPlan(hosts, scope, base, clone);
         if (values.force) note('--force is on: a conflicting real file moves to <name>.bak first.');
@@ -442,13 +614,13 @@ async function main() {
           ...(counts.skip > 0
             ? [`${counts.skip} name(s) skipped. Rename or remove the real files, then run install again.`]
             : []),
-          'next: run /setup-rstack in each host to configure models',
+          'next: run /setup-ramstack in each host to configure models',
         ]);
         return;
       }
       hosts ??= detectedOrAll();
       scope ??= 'global';
-      const base = resolveBase(scope, values['project-dir'] ?? saved?.projectDir);
+      const base = resolveBase(scope, values['project-dir']);
       if (!clone) syncStore();
       const counts = runLink(hosts, base, values.force, src);
       if (clone) configureGitHooks();
@@ -457,7 +629,7 @@ async function main() {
       if (counts.skip > 0) {
         say('note', `${counts.skip} name(s) skipped. Rename or remove the real files, then run install again.`);
       }
-      say('next', 'run /setup-rstack in each host to configure models');
+      say('next', 'run /setup-ramstack in each host to configure models');
       return;
     }
 
@@ -467,7 +639,7 @@ async function main() {
       const base = resolveBase(scope, values['project-dir'] ?? saved?.projectDir);
       if (wizard) {
         stageTotal = clone && !values['no-pull'] ? 3 : 2;
-        await banner('rstack update');
+        await banner('ramstack update');
         stage('Review');
         showPlan(hosts, scope, base, clone);
         note(
@@ -507,7 +679,7 @@ async function main() {
     const base = resolveBase(scope, values['project-dir'] ?? saved?.projectDir);
     if (wizard) {
       stageTotal = 2;
-      await banner('rstack uninstall');
+      await banner('ramstack uninstall');
       stage('Review');
       showPlan(hosts, scope, base, clone);
       note(
@@ -520,7 +692,7 @@ async function main() {
       runUnlink(hosts, base, src);
       finish(`links removed for ${hosts.join(', ')} (${scope})`, [
         clone
-          ? 'the clone, .rstack/config.local.json, and core.hooksPath stay'
+          ? 'the clone, .ramstack/config.local.json, and core.hooksPath stay'
           : `the store at ${STORE} stays; delete it to remove the copied skills`,
       ]);
       return;
@@ -528,11 +700,10 @@ async function main() {
     runUnlink(hosts, base, src);
     say('done', `links removed for ${hosts.join(', ')} (${scope})`);
   } catch (error) {
-    // Ctrl+D at a prompt rejects with AbortError; Ctrl+C never reaches here.
-    if (error?.code === 'ABORT_ERR') fail('\nerror: cancelled.');
+    restoreTty();
     throw error;
   } finally {
-    rl?.close();
+    stopWizardIo();
   }
 }
 

package/package.json

@@ -1,8 +1,8 @@
 {
   "name": "ramstack",
-  "version": "0.1.0",
+  "version": "0.2.0",
   "description": "Agent skills for Claude Code, Codex, and pi",
-  "author": "Justin Ramos <justinramos@outlook.com>",
+  "author": "Justin Ramos",
   "license": "MIT",
   "repository": {
     "type": "git",
@@ -16,7 +16,7 @@
     "skills"
   ],
   "bin": {
-    "rstack": "cli/rstack.mjs"
+    "ramstack": "cli/ramstack.mjs"
   },
   "engines": {
     "node": ">=18"

package/skills/arena/SKILL.md

@@ -25,7 +25,7 @@ The N candidates will receive the same prompt, so the prompt is the contract. Ge
 
 1. State the artifact each candidate is producing.
 2. Derive the rubric. State what success looks like for *this* task, then turn it into 3-6 concrete gradeable criteria. Concrete: `Adds a --dry-run flag that skips writes`. Vague: `code is correct`. The rubric is the picker's tool in Phase D; candidates only see the task.
-3. Pick the runners. Use `arena runners` from your host's `## <host>` section of `~/.agents/rstack-models.md` when present. Otherwise default to one each on `fable`, `opus`, `sonnet`. Spawn more when the arena covers multiple design directions. Same model N times when the work is generation-bound rather than judgment-sensitive.
+3. Pick the runners. Use `arena runners` from your host's `## <host>` section of `~/.agents/ramstack-models.md` when present. Otherwise default to one each on `fable`, `opus`, `sonnet`. Spawn more when the arena covers multiple design directions. Same model N times when the work is generation-bound rather than judgment-sensitive.
 4. Assign output paths. Each candidate writes to its own location (a git worktree where possible, otherwise `/tmp/arena-<slug>/candidate-<n>/`). N candidates writing to the same path is shared mutable state and fails the the **separate-before-serializing-shared-state** principle skill test.
 
 ## Phase B: Fan out
@@ -38,7 +38,7 @@ If a candidate fails to produce output, proceed with N-1 and note the dropout in
 
 ## Phase C: Cross-judge
 
-After all Phase B candidates complete, choose one model from the `arena cross-judge pool` in your host's `## <host>` section of `~/.agents/rstack-models.md` when present. Otherwise use `fable`, `opus`, `sonnet`. Prefer a different model family from the parent's. Spawn one readonly judge subagent on that model. It sees the rubric and the candidates by path label, scores each criterion, and recommends a base with rationale. It runs in parallel with the parent's reading in Phase D, not with the candidates themselves. Spawning while candidates are still writing means the judge sees partial or empty outputs and reports them as dropouts.
+After all Phase B candidates complete, choose one model from the `arena cross-judge pool` in your host's `## <host>` section of `~/.agents/ramstack-models.md` when present. Otherwise use `fable`, `opus`, `sonnet`. Prefer a different model family from the parent's. Spawn one readonly judge subagent on that model. It sees the rubric and the candidates by path label, scores each criterion, and recommends a base with rationale. It runs in parallel with the parent's reading in Phase D, not with the candidates themselves. Spawning while candidates are still writing means the judge sees partial or empty outputs and reports them as dropouts.
 
 ## Phase D: Pick a base
 

package/skills/automate-me/README.md

@@ -14,7 +14,7 @@ A `<handle>-mode/SKILL.md` skill file that encodes your response preferences, de
 
 ## How it works
 
-The skill mines recent agent transcripts in parallel slices for recurring patterns, asks you structured questions to fill gaps the mining missed, clusters findings into sections, drafts the skill (using the host's skill-authoring skill when available), and iterates on prose with the `unslop` skill. Patterns need 2+ independent sightings before they get codified. The `poteto-mode` skill serves as the reference shape.
+The skill mines recent agent transcripts in parallel slices for recurring patterns, asks you structured questions to fill gaps the mining missed, clusters findings into sections, drafts the skill (using the host's skill-authoring skill when available), and iterates on prose with the `unslop` skill. Patterns need 2+ independent sightings before they get codified. The `ram-mode` skill serves as the reference shape.
 
 ## Related
 

package/skills/automate-me/SKILL.md

@@ -60,7 +60,7 @@ Group the combined signals into sections. Common ones (use only what applies):
 - **Process**: git worktrees, commits, PRs, review/merge tooling.
 - **Skills**: skill-authoring habits, fix-the-skill-first, proposing new skills.
 
-The **poteto-mode** skill shows the shape. Read it for granularity. Don't copy its content; the user's rules are not the same as poteto-mode's.
+The **ram-mode** skill shows the shape. Read it for granularity. Don't copy its content; the user's rules are not the same as ram-mode's.
 
 ### 4. Draft the skill
 
@@ -104,6 +104,6 @@ Run a description-optimization loop only if the skill's trigger accuracy turns o
 
 ## Reference files
 
-- The **poteto-mode** skill: example of the output shape.
+- The **ram-mode** skill: example of the output shape.
 - The **unslop** skill: prose discipline for every line.
 - The host's skill-authoring skill (when installed): skill authoring process and writing guidelines.

package/skills/figure-it-out/SKILL.md

@@ -12,7 +12,7 @@ Don't reinvent a playbook you already have. A focused single-unit task that matc
 
 ## Start
 
-Open a todolist whose first item is to read the Principles section of the **poteto-mode** skill. Then add the phases below as todos.
+Open a todolist whose first item is to read the Principles section of the **ram-mode** skill. Then add the phases below as todos.
 
 ## Phase A: Frame
 

package/skills/interrogate/SKILL.md

@@ -33,7 +33,7 @@ Write one clear paragraph. Reviewers challenge whether the work achieves the int
 
 ## Step 3, Spawn Reviewers
 
-Launch all reviewers in a single message using the subagent tool. Use the `interrogate reviewers` list from your host's `## <host>` section of `~/.agents/rstack-models.md` when present, one reviewer per entry, extending or shrinking the Reviewer A/B/C labels below to the configured entry count; otherwise use the table defaults.
+Launch all reviewers in a single message using the subagent tool. Use the `interrogate reviewers` list from your host's `## <host>` section of `~/.agents/ramstack-models.md` when present, one reviewer per entry, extending or shrinking the Reviewer A/B/C labels below to the configured entry count; otherwise use the table defaults.
 
 | Subagent | Default model |
 |----------|---------------|

package/skills/{poteto-mode → ram-mode}/README.md

@@ -1,10 +1,10 @@
-# Poteto mode
+# Ram mode
 
 Opt-in rigor wrapper. Matches a task to a playbook, cites engineering principles, routes bulk work to subagents, and keeps the parent agent in the lead.
 
 ## When to use
 
-- `/poteto-mode` or "poteto"
+- `/ram-mode` or "ram-mode"
 - Any task that needs structured rigor: features, bug fixes, investigations, refactoring, shipping
 
 ## What it produces
@@ -13,7 +13,7 @@ A todolist driven by the matched playbook's steps, principle citations tied to r
 
 ## How it works
 
-The mode matches the task to one of 23 bundled playbooks (feature, bug fix, investigation, refactoring, and others). The playbook's steps become the todolist verbatim. Principles from the index are read as leaf skills and cited only when they change a real choice. Code-writing steps go to `poteto-agent` subagents. Review panels use the configured models. The parent reviews every diff and writes its own summary.
+The mode matches the task to one of 23 bundled playbooks (feature, bug fix, investigation, refactoring, and others). The playbook's steps become the todolist verbatim. Principles from the index are read as leaf skills and cited only when they change a real choice. Code-writing steps go to `ram-agent` subagents. Review panels use the configured models. The parent reviews every diff and writes its own summary.
 
 The mode is opt-in. `disable-model-invocation: true` prevents the host from applying it to prompts that just look similar.
 
@@ -22,5 +22,5 @@ The mode is opt-in. `disable-model-invocation: true` prevents the host from appl
 - [SKILL.md](SKILL.md)
 - [playbooks/](playbooks/) (23 playbook files)
 - [references/](references/) (bugbot triage, plan template)
-- [About poteto-mode](../../docs/explanation/about-poteto-mode.md)
+- [About ram-mode](../../docs/explanation/about-ram-mode.md)
 - [Playbooks reference](../../docs/reference/playbooks.md)

package/skills/{poteto-mode → ram-mode}/SKILL.md

@@ -1,14 +1,14 @@
 ---
-name: poteto-mode
-description: poteto's agent style for concise, detailed responses, deliberate subagents, unslopped prose, simple code, and verified work. Use for poteto, /poteto-mode, or requests to work in this style.
+name: ram-mode
+description: Ram-mode agent style for concise, detailed responses, deliberate subagents, unslopped prose, simple code, and verified work. Use for ram-mode, /ram-mode, or requests to work in this style.
 disable-model-invocation: true
 mode: true
 icon: crown
 color: yellow
-reminder: New task? Playbook match or rigor needed -> apply /poteto-mode. Casual turn or user opts out -> don't.
+reminder: New task? Playbook match or rigor needed -> apply /ram-mode. Casual turn or user opts out -> don't.
 ---
 
-# Poteto mode
+# Ram mode
 
 ## Non-negotiables
 
@@ -86,9 +86,9 @@ Read the leaf skill in full for any principle you apply. Each entry names when i
 
 ## Subagents
 
-**Use `subagent_type: "poteto-agent"` for any subagent you spawn inside a playbook step** (code-writing delegates, ad-hoc helpers). `/poteto-mode` and `poteto-agent` route through the same wrapper. Routed workflow skills (`how`, `why`, `interrogate`, `reflect`, `swarm`) set their own `subagent_type` for diverse-model review; respect what the skill prescribes, don't override to `poteto-agent`.
+**Use `subagent_type: "ram-agent"` for any subagent you spawn inside a playbook step** (code-writing delegates, ad-hoc helpers). `/ram-mode` and `ram-agent` route through the same wrapper. Routed workflow skills (`how`, `why`, `interrogate`, `reflect`, `swarm`) set their own `subagent_type` for diverse-model review; respect what the skill prescribes, don't override to `ram-agent`.
 
-**Defaults for every subagent call.** Run in the background when the host supports it, file pointers not inlined context, explicit model per role (configurable via `/setup-rstack`; defaults `sonnet` for code, `fable` for prose and judgment). Code delegates tier by difficulty. The hardest changes (cross-cutting design, gnarly concurrency, subtle algorithms) go to your strongest judgment model (`fable`) when the task needs judgment or the intent is vague, and to your strongest instruction-following model (`opus`) when the work is a precisely specified sequence of steps to execute to the letter; trivial mechanical edits go to your fast code model. Per-role lines in your host's section of the `/setup-rstack` config (`~/.agents/rstack-models.md`) override these defaults and the model choices in the routed skills (`how`, `why`, `arena`, `swarm`, `architect`, `interrogate`, `reflect`); a role with no line keeps its default, and a role line of `inherit-parent` or `auto` runs that role on the parent chat model (omit the subagent `model` param).
+**Defaults for every subagent call.** Run in the background when the host supports it, file pointers not inlined context, explicit model per role (configurable via `/setup-ramstack`; defaults `sonnet` for code, `fable` for prose and judgment). Code delegates tier by difficulty. The hardest changes (cross-cutting design, gnarly concurrency, subtle algorithms) go to your strongest judgment model (`fable`) when the task needs judgment or the intent is vague, and to your strongest instruction-following model (`opus`) when the work is a precisely specified sequence of steps to execute to the letter; trivial mechanical edits go to your fast code model. Per-role lines in your host's section of the `/setup-ramstack` config (`~/.agents/ramstack-models.md`) override these defaults and the model choices in the routed skills (`how`, `why`, `arena`, `swarm`, `architect`, `interrogate`, `reflect`); a role with no line keeps its default, and a role line of `inherit-parent` or `auto` runs that role on the parent chat model (omit the subagent `model` param).
 
 You own every subagent's work. Review the diff and write your own summary, don't pass through what it said. Interrupt-chained resumes silently drop directives, so fire a fresh subagent with consolidated scope rather than trusting a "done" summary. A second opinion is the same prompt against a different model. Agreement is high-signal.
 

package/skills/ram-mode/agents/openai.yaml

@@ -0,0 +1,5 @@
+interface:
+  display_name: "Ram Mode"
+  short_description: "Rigorous ram-mode working style"
+policy:
+  allow_implicit_invocation: false

package/skills/{poteto-mode → ram-mode}/playbooks/autonomous-run.md

@@ -3,10 +3,10 @@
 **You own the exit condition. Define done, then drive to it without stopping.** For "going to bed" / "run until done" / "/loop until X".
 
 1. State the exit condition as a checkable predicate before the first iteration (tests green, repro fixed, all N PRs merged, pixel-diff zero). A vague goal stalls; a predicate lets you stop.
-2. Pick the wake mechanism using the host's `/loop` command when it has one (a built-in, not a rstack skill). An event to watch (CI, a merge, a ref advancing) gets a watcher subagent that wakes you on the event, with a long time-based heartbeat as fallback. No event gets a fixed-interval heartbeat sized to when the result is worth re-checking.
+2. Pick the wake mechanism using the host's `/loop` command when it has one (a built-in, not a ramstack skill). An event to watch (CI, a merge, a ref advancing) gets a watcher subagent that wakes you on the event, with a long time-based heartbeat as fallback. No event gets a fixed-interval heartbeat sized to when the result is worth re-checking.
 3. Each iteration makes the smallest change the evidence justifies, verifies it against the predicate, commits if it advanced, discards changes that didn't help. Belt-and-suspenders that "might help" gets reverted, not left to ride.
    Sequence the work via the **sequence-verifiable-units** principle skill, verifying each unit before the next instead of batching checks at the end.
-4. Mid-run discoveries are yours. Address broken skills, related bugs, flaky verifiers, review noise, tooling failures, orphaned follow-ups, and fixable drift yourself via poteto-mode. Put out-of-band fixes in their own PR. Do not park reversible work for the human or use `AskUserQuestion`. Surface only irreversible actions, genuine product or preference calls no experiment can settle, or a real dead end. Keep the predicate as the main drive, and return to it after each side fix.
+4. Mid-run discoveries are yours. Address broken skills, related bugs, flaky verifiers, review noise, tooling failures, orphaned follow-ups, and fixable drift yourself via ram-mode. Put out-of-band fixes in their own PR. Do not park reversible work for the human or use `AskUserQuestion`. Surface only irreversible actions, genuine product or preference calls no experiment can settle, or a real dead end. Keep the predicate as the main drive, and return to it after each side fix.
 5. Checkpoint every iteration via the **show-me-your-work** skill, a row for what changed and whether the predicate moved. A run with no trail can't be audited or resumed.
 6. Stop when the predicate is met. A plateau is not a stop, so keep going and pivot your approach to push past it. Surface a genuine dead end rather than spinning, and never relax the predicate to declare victory.
 

package/skills/{poteto-mode → ram-mode}/references/plan.md

@@ -1,6 +1,6 @@
 # Plan
 
-Produce a phased implementation plan grounded in the **Principles** section of the `poteto-mode` skill. The plan is the deliverable. Do not implement.
+Produce a phased implementation plan grounded in the **Principles** section of the `ram-mode` skill. The plan is the deliverable. Do not implement.
 
 Open a todolist with one item per step below.
 
@@ -12,7 +12,7 @@ Plan when the change spans three or more files, introduces architecture, has com
 
 ## 1. Re-read principles
 
-Read the **Principles** section of the `poteto-mode` skill end to end, and the leaf `principle-*` skills it indexes. The principles govern every plan decision; cross-link them.
+Read the **Principles** section of the `ram-mode` skill end to end, and the leaf `principle-*` skills it indexes. The principles govern every plan decision; cross-link them.
 
 ## 2. Scope and constraints
 
@@ -24,7 +24,7 @@ Resolve what is in scope vs explicitly out, technical or platform constraints, p
 
 Delegate codebase exploration (the **guard-the-context-window** principle skill).
 
-- Prefer `subagent_type: "poteto-agent"`. `general-purpose` is the fallback. Never use the built-in `plan` subagent_type; it ignores this skill.
+- Prefer `subagent_type: "ram-agent"`. `general-purpose` is the fallback. Never use the built-in `plan` subagent_type; it ignores this skill.
 - Pass `model:` explicitly per the configured roles (defaults `sonnet` for code, `fable` for judgment).
 
 Each explorer returns file pointers, conventions, dependencies, test infrastructure, and entry points. No inlined dumps.
@@ -92,7 +92,7 @@ For bug fixes, the loop is reproduce on the surface, fix, verify on the same sur
 
 ## 6. Implementation guidance
 
-In the overview, name which poteto-mode non-negotiables the implementer must apply, by name:
+In the overview, name which ram-mode non-negotiables the implementer must apply, by name:
 
 - the **how** skill over each unfamiliar subsystem before changing it.
 - the **interrogate** skill for adversarial review on contested designs before shipping.

package/skills/{poteto-mode → ram-mode}/scripts/bootstrap.ts

@@ -11,7 +11,7 @@ const commanderPackagePath = join(
 );
 const installKeyPath = join(
   nodeModulesDirectory,
-  ".poteto-mode-tools-install-key"
+  ".ram-mode-tools-install-key"
 );
 
 function currentInstallKey(): string {

package/skills/{poteto-mode → ram-mode}/scripts/bun.lock

@@ -3,7 +3,7 @@
   "configVersion": 1,
   "workspaces": {
     "": {
-      "name": "@cursor-skill/poteto-mode-tools",
+      "name": "@ramstack/ram-mode-tools",
       "dependencies": {
         "commander": "14.0.0",
       },

package/skills/{poteto-mode → ram-mode}/scripts/package.json

@@ -1,5 +1,5 @@
 {
-  "name": "@rstack/poteto-mode-tools",
+  "name": "@ramstack/ram-mode-tools",
   "private": true,
   "type": "module",
   "scripts": {

package/skills/{setup-rstack → setup-ramstack}/README.md

@@ -1,16 +1,16 @@
-# Setup rstack
+# Setup ramstack
 
-Write the current host's section of `~/.agents/rstack-models.md`, the config file that sets a model per role for rstack skills.
+Write the current host's section of `~/.agents/ramstack-models.md`, the config file that sets a model per role for ramstack skills.
 
 ## When to use
 
-- `/setup-rstack` or "configure rstack models"
-- After installing rstack on a new host
+- `/setup-ramstack` or "configure ramstack models"
+- After installing ramstack on a new host
 - After model availability changes
 
 ## What it produces
 
-A `## <host>` section in `~/.agents/rstack-models.md` with one line per role. Other host sections in the file stay unchanged. Creates the file with a header comment if it does not exist.
+A `## <host>` section in `~/.agents/ramstack-models.md` with one line per role. Other host sections in the file stay unchanged. Creates the file with a header comment if it does not exist.
 
 ## How it works
 

package/skills/{setup-rstack → setup-ramstack}/SKILL.md

@@ -1,11 +1,11 @@
 ---
-name: setup-rstack
-description: Configure which models rstack uses per role, for the host you're running in (Claude Code, Codex, or pi). Detects your available models and writes your host's section of the shared config file. Use for /setup-rstack, "configure rstack models", or changing rstack's model choices.
+name: setup-ramstack
+description: Configure which models ramstack uses per role, for the host you're running in (Claude Code, Codex, or pi). Detects your available models and writes your host's section of the shared config file. Use for /setup-ramstack, "configure ramstack models", or changing ramstack's model choices.
 ---
 
-# Setup rstack
+# Setup ramstack
 
-Write your host's section of `~/.agents/rstack-models.md`, a config file that sets rstack's model per role. The skills read it at run time and fall back to their inline defaults when a section or line is absent, so this is an override layer, not a requirement.
+Write your host's section of `~/.agents/ramstack-models.md`, a config file that sets ramstack's model per role. The skills read it at run time and fall back to their inline defaults when a section or line is absent, so this is an override layer, not a requirement.
 
 The file is shared across hosts on purpose. Model slugs differ per host (`sonnet` means something in Claude Code and nothing in Codex or pi), so the file holds one `## <host>` section per tool, and every skill reads only the section for the host it is running in. Someone who switches between Claude Code, Codex, and pi runs this skill once in each and the tools stop stepping on each other.
 
@@ -25,7 +25,7 @@ If the host has a models API or CLI that lists the user's entitled models, prefe
 
 ### 2. Load current state
 
-If `~/.agents/rstack-models.md` exists and has a section for this host, read it and treat its values as the current choices. Otherwise start from the defaults in step 5. Never treat another host's section as current state; its slugs are not valid here.
+If `~/.agents/ramstack-models.md` exists and has a section for this host, read it and treat its values as the current choices. Otherwise start from the defaults in step 5. Never treat another host's section as current state; its slugs are not valid here.
 
 ### 3. Map and confirm
 
@@ -37,12 +37,12 @@ Every real slug written must be in the detected set; `inherit-parent` and `auto`
 
 ### 5. Write the config
 
-Rewrite this host's `## <host>` section of `~/.agents/rstack-models.md` with one line per role, using the same labels poteto-mode uses. Preserve every other host's section byte for byte; create the file with the header comment below if it doesn't exist. Re-runs stay idempotent by replacing the whole section.
+Rewrite this host's `## <host>` section of `~/.agents/ramstack-models.md` with one line per role, using the same labels ram-mode uses. Preserve every other host's section byte for byte; create the file with the header comment below if it doesn't exist. Re-runs stay idempotent by replacing the whole section.
 
 Shape, with the Claude Code defaults filled in (another host substitutes its own detected slugs under its own heading, keeping the same roles):
 
 ```
-# rstack model configuration. One `## <host>` section per tool; skills read only
+# ramstack model configuration. One `## <host>` section per tool; skills read only
 # the section for the host they run in. One line per role. Delete a line (or a
 # section) to fall back to the skill defaults.
 # `inherit-parent` or `auto` as a value: the role runs on the parent chat model
@@ -74,8 +74,8 @@ The role split follows the upstream design: fast mechanical code goes to the fas
 
 ### 6. Confirm
 
-Tell the user which host section was written and that skills read it on their next run. If the file has sections for other hosts, say so; if it doesn't, suggest running `/setup-rstack` inside each other tool they use so every host gets valid slugs.
+Tell the user which host section was written and that skills read it on their next run. If the file has sections for other hosts, say so; if it doesn't, suggest running `/setup-ramstack` inside each other tool they use so every host gets valid slugs.
 
 ### 7. Offer a verification skill (optional)
 
-Check whether the project has a way to drive the real app for proof (a `verify-*` skill, or an existing harness). If not, offer once: "want a project-local verification skill, so agents can drive the app the way a user does and prove changes work? I can generate one with /create-verification-skill." On yes, invoke `/create-verification-skill` (resolves wherever rstack is installed — workspace, user, or plugin). On no, move on without pushing.
+Check whether the project has a way to drive the real app for proof (a `verify-*` skill, or an existing harness). If not, offer once: "want a project-local verification skill, so agents can drive the app the way a user does and prove changes work? I can generate one with /create-verification-skill." On yes, invoke `/create-verification-skill` (resolves wherever ramstack is installed — workspace, user, or plugin). On no, move on without pushing.

package/skills/setup-ramstack/agents/openai.yaml

@@ -0,0 +1,3 @@
+interface:
+  display_name: "Setup ramstack"
+  short_description: "Configure ramstack models per host"

package/skills/swarm/SKILL.md

@@ -22,7 +22,7 @@ Open a todolist with one entry per phase before launching anything.
 1. State the done predicate and the artifact or report the swarm must return.
 2. Choose the shape. Partition into slices, race N workers on identical briefs, or mix both. For a race or mixed shape, declare `first pass`, `rank all`, or `best-of` before spawning.
 3. Set N from the user or derive it from the shape. N is total workers, not the cloud concurrency limit.
-4. Pick the worker model from `swarm workers` in your host's `## <host>` section of `~/.agents/rstack-models.md` when present. Otherwise use `sonnet`. For a model race, name each arm's model up front.
+4. Pick the worker model from `swarm workers` in your host's `## <host>` section of `~/.agents/ramstack-models.md` when present. Otherwise use `sonnet`. For a model race, name each arm's model up front.
 5. Give each worker its own writable output when it writes. Use a worktree, branch, or `/tmp/swarm-<slug>/worker-<n>/`.
 
 ## Phase B: Fan out

package/agents/poteto-agent.md

@@ -1,9 +0,0 @@
----
-name: poteto-agent
-description: Routing target for `/poteto-mode` and any request for poteto's style. Resume an existing `poteto-agent` for the conversation rather than spawning a sibling. Reads the `poteto-mode` skill's `SKILL.md` in full before any work, including its inline Principles index. Substituting `general-purpose` skips that read and drifts.
-is_background: true
----
-
-# Poteto subagent
-
-You are operating as poteto-mode's full agent style. Read the `poteto-mode` skill's `SKILL.md` in full before doing any work, including its inline Principles index. Navigate to a leaf `principle-*` skill whenever you apply that principle.

package/skills/poteto-mode/agents/openai.yaml

@@ -1,5 +0,0 @@
-interface:
-  display_name: "Poteto Mode"
-  short_description: "poteto's rigorous agent working style"
-policy:
-  allow_implicit_invocation: false

package/skills/setup-rstack/agents/openai.yaml

@@ -1,3 +0,0 @@
-interface:
-  display_name: "Setup rstack"
-  short_description: "Configure rstack models per host"